DESCRIPTION: (Adapted from applicant's description) Random digit dialing is becoming the standard strategy for selecting population controls in case- control studies conducted in the United States.The validity of this strategy, however, is difficult to assess primarily because the participation rate cannot be evaluated easily. The applicant will evaluate the validity of two random digit dialing strategies and of an alternative control group selection strategy in a pilot case-control study of breast cancer to be carried out in a population that is completely enumerated (i.e., the participation rate can be easily assessed). The study population consists of women who live in Jefferson County, Alabama and who have a Department of Motor Vehicles or a Health Care Financing Administration record. One hundred incident breast cancer cases in the study population will be identified using a network of local hospital tumor registries. A stratified sample of women will be selected from the study population and randomized into one of three groups, each corresponding to a control selection strategy. In strategy one, each potential control will be contacted as if her telephone number had been randomly generated. In strategy two, each potential control will be contacted as if her telephone number had been randomly generated and the name of the telephone subscriber obtained from directory assistance. In strategy three, each potential control will be sent a personal letter of presentation prior to telephone contact. A brief telephone interview will be used to collect information on established risk factors for breast cancer and other life-style factors. Overall participation rates will be documented for each control group. The frequency of specific reasons for non-participation, such as no contact and refusal will aslo be assessed. Odds ratios for each risk factor according to control group will be computed to evaluate the relative validity of the three control selection strategies. This investigation will provide empirical information on the validity of using random digit dialing and will evaluate alternative control group selection strategies.